论文部分内容阅读
为了提高考生考试行为识别的准确率,提出了一种视频监控中的考生异常行为识别方法ICanny-ABC-SVM。该算法从视频监控中提取考生行为图像,采用改进Canny算子对图像进行边缘检测;通过提取图像的不变矩特征,并将特征向量输入人工蜂群优化支持向量机中进行学习,构建考生行为分类器;运用仿真实验测试方法的性能。测试结果表明,此方法获得了较高的考生行为识别准确率与较快的识别速度,是一个性能较优的智能视频监控考生行为识别方法。
In order to improve the accuracy of the examinee’s behavior recognition, this paper proposes an ICanny-ABC-SVM method for identifying abnormal behavior of examinees in video surveillance. The algorithm extracts the candidate behavior image from the video surveillance, and adopts the improved Canny operator to detect the edge of the image. By extracting the moment invariant feature of the image and inputting the feature vector into the artificial bee swarm optimization support vector machine, Classifier; use simulation experiment to test the performance of the method. The test results show that this method obtains a higher recognition rate of examinee behavior recognition and faster recognition speed, which is a better performance of intelligent video monitoring candidate behavior recognition method.